Predictive Analytics on Diabetes Data using Machine Learning Techniques

Aim:

       To help doctors and practitioners in early prediction of diabetes using machine learning techniques.  

Abstract:

         Diabetes caused due to increase in amount of sugar or glucose which is condensed into the blood Identifying process of diabetes is the glucose and sugar levels needs to be checked before and after meal, there are fluctuations before and after meal, this whole process of patient visiting a doctor is tiresome. But in Machine Learning algorithms helps us to solve this issue. The motive of this study and research is to make use of features and to predict the likelihood of the disease, Decision Tree, Random Forest, K Nearest Neighbours, Naive Bayes and Support Vector Machine are the algorithm that have been applied to detect and predict diabetes at an early stage. A dataset of a patient’s medical record is obtained and different machine learning algorithms are applied on the dataset. Performance and accuracy of the applied algorithms is discussed and compared. Comparison of the different machine learning techniques used in this study reveals which algorithm is best suited for prediction of diabetes.

Existing System:

           It is apparent from the truth that the occurrence of diabetes mellitus is high and the complication in the prevention of diabetes also increases. Thus, there are many patients who need the required knowledge and skills to enrich their health. In such cases, the patients are needed to visit the diagnostic center for their treatment. Because of this, they lost their time and expenses. The data of diabetic patients collected from the UCI laboratory is used to discover patterns with KNN and Support Vector Machine (SVM). The results are compared for performance and accuracy with these algorithms. It gives some moderate level of accuracy .So we use more ML algorithms to predict diabetic accurately using web framework.

Proposed System:

            In this paper, we are using machine learning algorithms to predict diabetes disease. The client on their first login has to register themselves on the Web Application. The web Application created by Django. Once the user logins into the system he gets all the access for predicting the diabetic and by  using the input such as Pregnancies, Glucose, Blood Pressure, Skin Thickness, BMI, insulin level and age based on their own. After submitting the inputs, it’s move on to the trained model for comparison. Already trained model were trained by machine learning algorithms. Algorithms used for training a dataset are K Nearest Neighbours (KNN), Naive Bayes (NB), Support Vector Machine (SVM), Decision Tree (DT) and Random Forest (RF). Comparison of the different machine learning techniques used in this study reveals which algorithm is best suited for prediction of diabetes.